INDEX

Explanations

phrases related to consequences or outcomes

oai_token-act-pair · gpt-3.5-turbo

phrases indicating personal struggles or obstacles

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 11-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.11.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.11.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

Individual

-0.62

Prop

-0.58

ourse

-0.57

Federal

-0.57

andum

-0.55

 Associated

-0.55

National

-0.55

Information

-0.54

According

-0.54

 Individuals

-0.53

POSITIVE LOGITS

 dont

1.08

 didnt

1.05

 doesnt

1.04

 kinda

1.02

 alot

0.97

 gotta

0.96

 fucked

0.93

 REALLY

0.91

 wont

0.88

 shit

0.86

Activations Density 0.936%

phrases related to consequences or outcomes

phrases indicating personal struggles or obstacles

No Comments

No Known Activations