INDEX

Explanations

phrases related to hierarchical structures or organizations

oai_token-act-pair · gpt-3.5-turbo

occurrences of the term "ranks" 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

argo

-0.80

issance

-0.76

 Robo

-0.63

 Breath

-0.58

 refere

-0.57

Deal

-0.56

 Monstrous

-0.56

positive

-0.56

 Archangel

-0.56

 Pilgrim

-0.56

POSITIVE LOGITS

 ranks

0.98

enance

0.84

 insign

0.81

chwitz

0.77

staff

0.75

leading

0.74

 rank

0.74

Rank

0.74

ename

0.73

chel

0.72

Activations Density 0.011%

phrases related to hierarchical structures or organizations

occurrences of the term "ranks" in various contexts

No Comments

No Known Activations