INDEX

Explanations

verbs related to actions or activities

oai_token-act-pair · gpt-3.5-turbo

action verbs that indicate engagement or activity

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 4-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.4.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.4.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

Applic

-0.78

FTWARE

-0.75

CLR

-0.64

 Apart

-0.62

adh

-0.62

pload

-0.61

 unforeseen

-0.59

 Garage

-0.59

HCR

-0.58

 Administration

-0.57

POSITIVE LOGITS

poke

0.73

idas

0.67

backs

0.67

ida

0.66

adelphia

0.65

 ourselves

0.64

 caucuses

0.64

bol

0.64

birds

0.63

saw

0.62

Activations Density 0.383%

verbs related to actions or activities

action verbs that indicate engagement or activity

No Comments

No Known Activations