INDEX

Explanations

phrases related to the name "Gan."

oai_token-act-pair · gpt-3.5-turbo

repeated mentions of specific names or entities

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

 Wolves

-0.68

ifications

-0.68

 rack

-0.61

itud

-0.61

 Rapt

-0.60

 cabinets

-0.59

ition

-0.59

 bills

-0.58

inct

-0.58

 sockets

-0.58

POSITIVE LOGITS

pour

0.92

Gan

0.92

esh

0.89

elist

0.80

bare

0.78

zu

0.77

noon

0.77

zeb

0.76

Ò

0.75

roo

0.75

Activations Density 0.041%

phrases related to the name "Gan."

repeated mentions of specific names or entities

No Comments

No Known Activations