INDEX

Explanations

adjectives describing the strength or intensity of actions or states

oai_token-act-pair · gpt-3.5-turbo

references to governance or procedural issues

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 8-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.8.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.8.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

 behavi

-0.77

ulence

-0.74

 tremend

-0.67

reality

-0.65

lihood

-0.63

 ethics

-0.63

figure

-0.62

 neigh

-0.62

NESS

-0.62

 relent

-0.62

POSITIVE LOGITS

pleted

1.15

 Played

1.12

ired

1.12

iked

1.07

ried

1.07

arthed

1.05

asted

1.04

Used

1.03

 Bought

1.02

Posted

1.01

Activations Density 0.229%

adjectives describing the strength or intensity of actions or states

references to governance or procedural issues

No Comments

No Known Activations