INDEX

Explanations

high-level actions or concepts related to large-scale operations

oai_token-act-pair · gpt-3.5-turbo

references to large scale and high-level concepts or operations

oai_token-act-pair · gpt-4o-mini Triggered by @bot

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 7-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.7.hook_resid_pre

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

Skylion007/openwebtext

Features

24,576

Data Type

torch.float32

Hook Point

blocks.7.hook_resid_pre

Architecture

standard

Context Size

128

Dataset

Skylion007/openwebtext

Hook Point Layer

Activation Function

relu

Embeds

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

adelphia

-0.76

oother

-0.73

ename

-0.73

ÃĥÃĤÃĥÃĤÃĥÃĤÃĥÃĤÃĥÃĤÃĥÃĤÃĥÃĤÃĥÃĤÃĥÃĤÃĥÃĤÃĥÃĤÃĥÃĤÃĥÃĤÃĥÃĤÃĥÃĤÃĥÃĤÃĥÃĤÃĥÃĤÃĥÃĤÃĥÃĤÃĥÃĤÃĥÃĤÃĥÃĤÃĥÃĤÃĥÃĤÃĥÃĤÃĥÃĤÃĥÃĤÃĥÃĤÃĥÃĤÃĥÃĤÃĥÃĤ

-0.73

bright

-0.72

quet

-0.71

stars

-0.71

TOP

-0.70

 Cosponsors

-0.70

 Clarks

-0.68

POSITIVE LOGITS

 inspection

0.83

 involvement

0.80

 interaction

0.80

 development

0.78

 diplomatic

0.75

 confrontation

0.74

 intervention

0.74

 reconnaissance

0.74

 cooperation

0.73

 administrative

0.73

Activations Density 0.088%

high-level actions or concepts related to large-scale operations

references to large scale and high-level concepts or operations

No Comments

No Known Activations