INDEX

Explanations

phrases referring to direct actions or relationships

oai_token-act-pair · gpt-3.5-turbo

phrases related to direct communication or actions

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

etheless

-0.77

 cautiously

-0.76

omaly

-0.75

bryce

-0.66

 brightest

-0.63

pse

-0.63

conserv

-0.63

manship

-0.62

dust

-0.60

 inconsist

-0.60

POSITIVE LOGITS

achable

0.84

avez

0.64

OURCE

0.64

SpaceEngineers

0.64

iration

0.64

ledged

0.63

 Nieto

0.63

 Manufacturer

0.62

ividual

0.61

heny

0.61

Activations Density 0.167%

phrases referring to direct actions or relationships

phrases related to direct communication or actions

No Comments

No Known Activations