INDEX

Explanations

terms related to boycotts

oai_token-act-pair · gpt-3.5-turbo

references to boycotts

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 2-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.2.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.2.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

 confir

-1.13

 reluct

-1.10

 exha

-0.89

 ingred

-0.85

 millenn

-0.84

ortunately

-0.84

 tremend

-0.82

 actu

-0.81

senal

-0.81

 unnecess

-0.80

POSITIVE LOGITS

cott

1.47

cot

0.91

age

0.89

andise

0.85

ijn

0.85

ages

0.84

AFB

0.84

arer

0.84

arget

0.83

reath

0.83

Activations Density 0.008%

terms related to boycotts

references to boycotts

No Comments

No Known Activations