INDEX

Explanations

words related to legal or procedural terminology

oai_token-act-pair · gpt-3.5-turbo

past tense verbs in a variety of contexts

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 5-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.5.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.5.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

isha

-0.73

 contrace

-0.70

ishi

-0.69

 foremost

-0.68

izont

-0.67

jew

-0.63

asha

-0.63

 Belg

-0.62

 motto

-0.61

ISH

-0.61

POSITIVE LOGITS

umbnail

0.98

ragon

0.92

ures

0.85

enced

0.82

vous

0.81

ornia

0.81

ure

0.74

uits

0.74

inct

0.74

Activations Density 0.028%

words related to legal or procedural terminology

past tense verbs in a variety of contexts

No Comments

No Known Activations