INDEX

Explanations

proper names of individuals

oai_token-act-pair · gpt-3.5-turbo

different phrases or elements related to legal and official contexts

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

ishers

-0.77

ãĥ¥

-0.67

rolley

-0.61

acters

-0.60

eals

-0.58

rats

-0.58

vable

-0.58

psy

-0.57

olars

-0.56

 pursuit

-0.56

POSITIVE LOGITS

Jr

1.17

PhD

0.97

aka

0.91

Sr

0.82

who

0.76

 resigned

0.75

Jr

0.75

LLP

0.74

anu

0.70

 intervened

0.68

Activations Density 0.111%

proper names of individuals

different phrases or elements related to legal and official contexts

No Comments

No Known Activations