INDEX

Explanations

phrases related to conditions set by certain criteria or standards

oai_token-act-pair · gpt-3.5-turbo

conditional statements and potential outcomes

oai_token-act-pair · gpt-4o-mini Triggered by @bot

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 8-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.8.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.8.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

ajor

-0.56

 Hels

-0.55

ãĤ»

-0.55

çīĪ

-0.55

 Chak

-0.53

GN

-0.53

 paternal

-0.52

senal

-0.52

 looph

-0.51

bestos

-0.51

POSITIVE LOGITS

 then

0.92

 anymore

0.90

 THEN

0.88

 please

0.84

then

0.83

?,

0.79

 thats

0.76

 Then

0.74

 beforehand

0.73

 PLEASE

0.71

Activations Density 0.779%

phrases related to conditions set by certain criteria or standards

conditional statements and potential outcomes

No Comments

No Known Activations