INDEX

Explanations

proper nouns, in this case specifically related to a political figure

oai_token-act-pair · gpt-3.5-turbo

mentions of the name "Mund" and its variations

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 3-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.3.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.3.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

ãĤ·ãĥ£

-0.82

-0.79

OWER

-0.71

renches

-0.70

olkien

-0.68

ardless

-0.67

chnology

-0.66

MET

-0.66

OHN

-0.66

OPLE

-0.65

POSITIVE LOGITS

ele

0.84

nih

0.83

ias

0.83

 Mund

0.81

0.77

us

0.76

lock

0.76

ia

0.75

los

0.75

ality

0.74

Activations Density 0.005%

proper nouns, in this case specifically related to a political figure

mentions of the name "Mund" and its variations

No Comments

No Known Activations