INDEX

Explanations

phrases mentioning the word "needed."

oai_token-act-pair · gpt-3.5-turbo

instances of the word "needed."

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

imore

-0.77

vez

-0.75

izon

-0.72

gdala

-0.72

creen

-0.70

cow

-0.69

pret

-0.68

ipel

-0.66

eve

-0.64

ahar

-0.64

POSITIVE LOGITS

lessly

0.90

 urgently

0.74

 sshd

0.71

 introdu

0.70

nces

0.67

 unanim

0.66

 Earthqu

0.65

 reinforcements

0.65

 assistance

0.65

HAEL

0.64

Activations Density 0.031%

phrases mentioning the word "needed."

instances of the word "needed."

No Comments

No Known Activations