INDEX

Explanations

pairs of keywords associated with a technical or programming context

oai_token-act-pair · gpt-3.5-turbo

terms related to operators in a programming or mathematical context

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 3-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.3.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.3.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

theless

-1.02

BOOK

-0.74

USER

-0.73

forest

-0.69

 Caucus

-0.68

 Dakota

-0.65

 Shal

-0.63

Charge

-0.63

staking

-0.63

light

-0.61

POSITIVE LOGITS

 oper

1.07

ateg

1.04

ational

1.02

atically

1.01

atic

0.97

boa

0.91

andi

0.87

estate

0.87

aton

0.86

arie

0.84

Activations Density 0.014%

pairs of keywords associated with a technical or programming context

terms related to operators in a programming or mathematical context

No Comments

No Known Activations