INDEX

Explanations

words related to being open, available, and easily reached

oai_token-act-pair · gpt-3.5-turbo

references to accessibility 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 @ 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

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

ortium

-0.78

nance

-0.74

chief

-0.72

origin

-0.70

uke

-0.68

hunter

-0.67

protein

-0.66

 Origin

-0.66

hun

-0.66

ordon

-0.64

POSITIVE LOGITS

ibility

1.01

accessible

0.98

 accessible

0.94

ible

0.92

isable

0.82

izable

0.80

ibly

0.77

 adolesc

0.74

[*

0.72

ahime

0.69

Activations Density 0.011%

words related to being open, available, and easily reached

references to accessibility in various contexts

No Comments

No Known Activations