INDEX

Explanations

ways of thinking or phrases related to the concept of thinking

oai_token-act-pair · gpt-3.5-turbo

references to the concept of "thinking" 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 @ 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

Ann

-0.69

 earthqu

-0.65

Merit

-0.64

 tears

-0.63

lehem

-0.63

Pie

-0.63

anca

-0.62

feeding

-0.61

ocal

-0.61

 Adin

-0.61

POSITIVE LOGITS

 thinkers

0.85

 thinker

0.84

fulness

0.80

geist

0.80

lass

0.76

itar

0.74

LAB

0.73

 minds

0.73

 bubble

0.72

ortment

0.71

Activations Density 0.052%

ways of thinking or phrases related to the concept of thinking

references to the concept of "thinking" in various contexts

No Comments

No Known Activations