INDEX

Explanations

phrases that indicate statistical data or numerical percentages

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

New Auto-Interp

Configuration

jbloom/GPT2-Small-OAI-v5-128k-attn-out-SAEs/v5_128k_layer_0

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

Skylion007/openwebtext

Features

131,072

Data Type

torch.float32

Hook Name

blocks.0.hook_attn_out

Hook Layer

Architecture

standard

Context Size

Dataset

Skylion007/openwebtext

Activation Function

topk

Embeds

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

sonian

-0.87

lass

-0.86

gow

-0.83

æ©

-0.82

wright

-0.81

hum

-0.79

agall

-0.78

stadt

-0.74

daq

-0.74

ther

-0.74

POSITIVE LOGITS

ãĤ»

0.71

 Leban

0.70

ndum

0.69

 Nights

0.68

Mex

0.67

abouts

0.67

ãĥ¼ãĥ³

0.66

Ago

0.65

 Baird

0.64

encies

0.63

Activations Density 0.338%

phrases that indicate statistical data or numerical percentages

No Comments

No Known Activations