INDEX

Explanations

words related to historical events and locations, especially related to wars and political movements

oai_token-act-pair · gpt-3.5-turbo

specific names or terms related to political entities or groups

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

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

 timet

-0.63

 margins

-0.62

 backgrounds

-0.60

gif

-0.59

 Attribution

-0.57

 trimmed

-0.57

\<

-0.56

ãĤ¦

-0.55

 barr

-0.54

 traged

-0.53

POSITIVE LOGITS

geon

0.89

hyde

0.85

alion

0.79

bag

0.79

step

0.75

osaurs

0.74

 Restaur

0.74

geons

0.73

othy

0.73

eteria

0.72

Activations Density 0.126%

words related to historical events and locations, especially related to wars and political movements

specific names or terms related to political entities or groups

No Comments

No Known Activations