INDEX

Explanations

instances where support is expressed towards a particular person or cause

oai_token-act-pair · gpt-3.5-turbo

references to political endorsements or backing

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

 Shant

-0.77

 earthqu

-0.68

oufl

-0.67

 pores

-0.66

Â¯Â¯

-0.65

ModLoader

-0.63

 Hole

-0.62

juries

-0.61

bum

-0.61

 Goddard

-0.61

POSITIVE LOGITS

votes

0.93

itism

0.89

Support

0.88

enance

0.81

vote

0.81

 endorse

0.81

byn

0.80

 endorsing

0.80

Supporters

0.77

 supporters

0.77

Activations Density 0.057%

instances where support is expressed towards a particular person or cause

references to political endorsements or backing

No Comments

No Known Activations