INDEX

Explanations

terms related to physical activities or actions

oai_token-act-pair · gpt-3.5-turbo

repeated mentions of the word "activity" across various contexts

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 1-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.1.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.1.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

 Missile

-0.69

ixed

-0.68

adra

-0.66

haul

-0.65

Nak

-0.64

uran

-0.63

 veiled

-0.63

oiler

-0.63

eros

-0.63

 passage

-0.62

POSITIVE LOGITS

 activity

1.17

 activities

0.93

activity

0.92

 Activity

0.92

ername

0.90

 newsp

0.85

uality

0.84

ivism

0.80

ual

0.79

ickr

0.78

Activations Density 0.012%

terms related to physical activities or actions

repeated mentions of the word "activity" across various contexts

No Comments

No Known Activations