INDEX

Explanations

phrases that contrast viewpoints or actions between different entities

oai_token-act-pair · gpt-3.5-turbo

references to "others" in various contexts

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 11-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.11.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.11.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

"},"

-0.82

Opening

-0.69

Pen

-0.68

Awesome

-0.65

Alright

-0.64

United

-0.64

SI

-0.63

Rated

-0.63

 Annotations

-0.62

RNA

-0.62

POSITIVE LOGITS

 merely

1.08

 simply

1.05

 prefer

0.94

 succumb

0.89

 remain

0.80

 cling

0.79

 rely

0.78

 just

0.78

are

0.77

opt

0.76

Activations Density 0.124%

phrases that contrast viewpoints or actions between different entities

references to "others" in various contexts

No Comments

No Known Activations