INDEX

Explanations

words related to motivation or the underlying reasons for actions

oai_token-act-pair · gpt-3.5-turbo

references to reasons behind actions or behaviors

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

esan

-0.84

opic

-0.76

hold

-0.75

 Honour

-0.74

ocker

-0.73

oys

-0.71

iannopoulos

-0.69

 Islands

-0.68

abb

-0.65

rooms

-0.65

POSITIVE LOGITS

ivated

1.04

 motivated

0.99

TextColor

0.91

ItemTracker

0.84

 motivate

0.84

 motivation

0.84

 motivations

0.82

ivation

0.80

 motivating

0.79

 compel

0.79

Activations Density 0.014%

words related to motivation or the underlying reasons for actions

references to reasons behind actions or behaviors

No Comments

No Known Activations