INDEX

Explanations

phrases related to official government actions or involvement

oai_token-act-pair · gpt-3.5-turbo

mentions of political events and actions involving government leaders

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 9-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.9.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.9.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

 percussion

-0.63

Vel

-0.63

UFC

-0.62

 trapping

-0.62

 capt

-0.62

 refine

-0.61

 Velocity

-0.61

riks

-0.60

 Takeru

-0.60

Zot

-0.60

POSITIVE LOGITS

 assassinated

0.95

 appoint

0.93

 presided

0.91

 impeachment

0.89

 speeches

0.83

 reelection

0.82

 veto

0.82

 appointing

0.80

cius

0.78

Tea

0.78

Activations Density 0.863%

phrases related to official government actions or involvement

mentions of political events and actions involving government leaders

No Comments

No Known Activations