INDEX

Explanations

proper nouns related to politics, economics, and classified information

oai_token-act-pair · gpt-3.5-turbo

references to specific individuals or names connected to significant events or honors

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 10-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.10.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.10.hook_resid_pre

Architecture

standard

Context Size

128

Dataset

Skylion007/openwebtext

Hook Point Layer

Activation Function

relu

Embeds

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

RD

-0.83

greg

-0.83

ral

-0.83

RG

-0.82

ull

-0.79

roth

-0.79

raltar

-0.79

RAD

-0.78

RO

-0.78

ril

-0.78

POSITIVE LOGITS

 Casey

0.76

USE

0.74

etts

0.73

ante

0.73

anted

0.72

ants

0.70

Pam

0.69

iment

0.68

Sue

0.67

Activations Density 0.397%

proper nouns related to politics, economics, and classified information

references to specific individuals or names connected to significant events or honors

No Comments

No Known Activations