INDEX

Explanations

mentions of political candidates, specifically presidential candidates

oai_token-act-pair · gpt-3.5-turbo

references to presidential candidates

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 5-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.5.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.5.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

ãĥĥ

-0.86

ãĤ®

-0.75

shapeshifter

-0.75

ãĥĥãĤ¯

-0.74

ateral

-0.72

ãĥ¥

-0.71

rex

-0.70

ãĥĥãĥī

-0.70

ãĥķãĤ¡

-0.70

elly

-0.69

POSITIVE LOGITS

Â¯

0.79

 Jacket

0.79

 Huck

0.70

 Hail

0.67

 Emmanuel

0.66

 Polly

0.66

 Narendra

0.66

Bid

0.64

INTON

0.63

ratulations

0.62

Activations Density 0.057%

mentions of political candidates, specifically presidential candidates

references to presidential candidates

No Comments

No Known Activations