INDEX

Explanations

words related to importance, impact, existence, extent, and role

oai_token-act-pair · gpt-3.5-turbo

references to significant issues and the extent of their impact

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

arna

-0.80

ANA

-0.68

INA

-0.67

ETA

-0.67

adish

-0.66

inka

-0.65

itches

-0.63

MpServer

-0.62

yne

-0.62

JP

-0.61

POSITIVE LOGITS

 extent

0.80

 validity

0.75

 dangers

0.71

lex

0.70

 viability

0.70

 inherent

0.69

 unfolding

0.69

 evolution

0.69

 ramifications

0.67

 nature

0.66

Activations Density 0.356%

words related to importance, impact, existence, extent, and role

references to significant issues and the extent of their impact

No Comments

No Known Activations