INDEX

Explanations

names or terms related to a person named "Hab."

oai_token-act-pair · gpt-3.5-turbo

references to specific individuals or groups associated with a particular context

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

ãĤ´ãĥ³

-0.74

afort

-0.71

aunder

-0.70

merce

-0.70

çĶ°

-0.68

NCT

-0.68

 reconstruction

-0.67

bilt

-0.65

senal

-0.65

 Piercing

-0.65

POSITIVE LOGITS

erer

0.95

Hab

0.91

erences

0.90

itual

0.87

erers

0.86

lar

0.83

pak

0.82

ility

0.81

ered

0.80

script

0.78

Activations Density 0.030%

names or terms related to a person named "Hab."

references to specific individuals or groups associated with a particular context

No Comments

No Known Activations