INDEX

Explanations

questions or statements asking for specific information or explanations

oai_token-act-pair · gpt-3.5-turbo

inquiries or phrases seeking clarification or emphasis on specific details

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

eka

-0.67

Ai

-0.62

cler

-0.61

entimes

-0.61

Pg

-0.61

meric

-0.60

pora

-0.60

Phi

-0.59

 gorilla

-0.58

ii

-0.58

POSITIVE LOGITS

irements

0.82

ACTION

0.65

abyte

0.63

LOG

0.63

Ùĩ

0.62

cellence

0.61

ILLE

0.61

LOAD

0.61

edIn

0.61

ãĤ»

0.59

Activations Density 0.097%

questions or statements asking for specific information or explanations

inquiries or phrases seeking clarification or emphasis on specific details

No Comments

No Known Activations