INDEX

Explanations

words related to projects or campaigns

oai_token-act-pair · gpt-3.5-turbo

repeated mentions of initiatives or projects

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

oil

-0.93

parts

-0.79

bane

-0.77

oiler

-0.77

ummies

-0.71

ibur

-0.65

oak

-0.65

Parts

-0.64

fell

-0.64

gone

-0.63

POSITIVE LOGITS

 initiative

1.08

eers

1.05

ually

0.81

 initiatives

0.79

ĸļ

0.78

 effort

0.77

 endeavor

0.77

interstitial

0.75

udence

0.72

 minded

0.72

Activations Density 0.012%

words related to projects or campaigns

repeated mentions of initiatives or projects

No Comments

No Known Activations