INDEX

Explanations

mentions of wells

oai_token-act-pair · gpt-3.5-turbo

references to the word "well" in various contexts

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

IENCE

-0.72

hip

-0.71

ategory

-0.67

iferation

-0.67

İĭ

-0.65

 blasphemy

-0.64

atto

-0.63

©¶æ¥µ

-0.62

ierrez

-0.62

anish

-0.61

POSITIVE LOGITS

spring

0.80

ards

0.80

esley

0.79

coat

0.78

enough

0.74

acent

0.73

orah

0.70

reads

0.70

 behaved

0.70

 suited

0.69

Activations Density 0.035%

mentions of wells

references to the word "well" in various contexts

No Comments

No Known Activations