INDEX

Explanations

descriptions or discussions about specific elements within a larger context

oai_token-act-pair · gpt-3.5-turbo

references to game design and improvements

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 7-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.7.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.7.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

 intervened

-0.68

govtrack

-0.65

apo

-0.64

 istg

-0.64

 Choice

-0.62

akening

-0.62

cession

-0.60

 ALWAYS

-0.60

ollar

-0.60

fing

-0.59

POSITIVE LOGITS

 similarities

1.17

 resemblance

1.05

 familiar

1.04

 characteristics

1.04

 resemb

1.00

 wrinkles

0.97

 themes

0.93

 tropes

0.92

 specs

0.92

 specifications

0.92

Activations Density 1.039%

descriptions or discussions about specific elements within a larger context

references to game design and improvements

No Comments

No Known Activations